I have just started working on open loop FOG, can you suggest few literature on it.
Which FOG are you investigating right now? What performance needs are driving you to consider them? For basic terms, test methods and performance metrics, I would probably start with IEEE-STD-952-1997. While I haven't read this entire paper, this looks like it provides some good details on operations and limits of the technology:
One more thing to think about, the ADIS16137 might worth looking at as well. I am not sure what your performance objectives are, but this product might give you some cost advantages if the performance meets your needs.
I hope that helps!
By the way, I am sorry that I didn't mention this in my first post but we typically try to start new discussions for new questions. That way, these discussions will not be too long for future users to review when they need help. Many thanks!
First of all I thank you for ur reply,
I am working on Fiber optic gyro, open loop and single axis phase modulated by a PZT.
±4000°/sec measurement range
ADIS16137 this seems to be a MEMs based gyro. I am looking for interferometer FOG.
I actually thought of mailing you but i couldn't find ur mail ID so continued the conversation.
U can mail me on firstname.lastname@example.org
As per our calculations Input power to my FOG electronics is 40uW to 65uW (5mW is the laser diode power, 200m PM fiber spool, 1Mhz PZT freq)
http://pubcouncil.kuniv.edu.kw/jer/files/07Nov201202070049-68_04_A%20novel.pdf as per this document do I need to capture signal frequency up to 4-5MHz, as per my resolution i need a 19bit ADC. To this specifications I could not find a ADC.
" A novel technique for spectral processing of a fiber optic gyroscope" for as per this document do I need to capture signal frequency up to 4-5MHz, as per my resolution i need a 19bit ADC. To this specifications I could not find a ADC.
Is this the paper that you are referring to?
Based on the key ADC specifications you have offered, it would appear like the AD7760 is the closest solution I can point you towards: www.analog.com/AD7760
I hope that helps.
Hello good evening NevadaMark,
Ya this is the paper we are referring (I tried to upload the link but there was some problem by which i could not do, so I have typed its name...)
Ya sir we also found this ADC to be OK, but its full power bandwidth is limited to 1.3Mhz. Because of this reason I could not meet my 4Mhz data acquiring...
Do I have any other possibility.
As I said in my last post, that is the ADC that is the closest to meeting the requirements you state. While I have not had time to read every part of the referenced paper, why not use the same approach that this paper uses, which leverages the AD7679 in an under-sampling approach?
Is there any approach to increase my 18 bit(AD7679) resolution to 20bits.
I might be missing something important, but won't the AD7760 provide this? The AD7679 provides 18-bits at 570 kSPS, while the AD7760 provides 24-bits at 2.5 MSPS. Admittedly, I haven't done an exhaustive analysis on this, but wanted to make sure I understood any limitations you had already noted.
I have gone through all the ADC makes, none of the device support my application.
ADI seems to be better option, as it is near to my requirement.
But is there any option/scheme which will help me to sought out this.
We appreciate that. That is one area that ADI has specialized in for quite some time. With that in mind, I think that we have already answered with the solution that is closest to meeting your stated needs (AD7760), but you last question was focused on improving the AD7679, from 18-bits to 20-bits of resolution. Perhaps I misunderstood, but I took that to mean that you were re-evaluating your needs, using the under sampling approach in the cited reference and had decided that better resolution, at the AD7679 sample rates, might be valuable in your present research. Did I misunderstand? If I understood correctly, does the AD7760 meet this objective?
I may not be knowing all the approaches/topologies for increasing resolution.
As ADI is expert in data converters I have posted a query.
I knew that by over sampling resolution can be increased, but with this ADC there is a no chance of sampling at nyquist rate.
I want to know one thing as my signal is bandpass signal, I can undersample it as per shannon criteria. If i sample it at higher rate then by averaging can I improve its resolution.
I mean to say that undersampling then oversampling.
Do we have any simulation tools for this purpose.
We appreciate the recognition of our ADC capability. One thing that has helped us grow that capability is starting with a clear definition of needs and if necessary, re-definition of these needs when the preliminary desires are not available. Just to re-cap, you asked for 19-bit, 4-5MHz sample rates. The AD7960 meets the sample rate requirements, but only offers 18-bit outputs. The AD7760 offers 24-bit resolution, but at 2.5MSPS. If I understand the oversampling technique, it has certain noise level and distribution requirements, with respect to the LSB weight and sample rates.
So, are you asking, "can you average AD7960 samples together, to improve its resolution to 19 bits? If not, can you correct this statement, but maintain its level of definition?
BTW, I didn't mention this, but this level of definition will help me identify if we have any simulation tools or estimation techniques that we can share.
Yes, do averaging samples improve resolution.
I did not understand "BTW"
I am sorry about the short-hand. "BTW" means "by the way." I sometimes forget that I am not sending a text to my 16-year old daughter. Sorry about that!
I am wondering if it would make sense to run 4 AD7960 ADCs in parallel and sum their outputs together, in order to improve the resolution by 2 bits. When you add the outputs together, the signal will increase by 4x, while the noise will only increase by a factor of 2.
It looks like my last sentence was deleted on accident. When the signal increases by a factor of 4 and the noise increases by a factor of 2, you can shift the binary number two bits to the right to "re-scale" the number and complete the average. When doing this, make sure you retain the bit growth by expanding the width of the number throughout your signal processing chain.
Retrieving data ...